Академический Документы
Профессиональный Документы
Культура Документы
Controlled Document
Project N/A
Document Revision 02
ECCN EAR 99
Disclosure None
Revision History shown on next
page
Revision History
REVISION STATUS APPROVAL
Rev. Date Description Originator Reviewer Approver
01 2011-11 Draft for Review Matt Madden Alistair Salisbury Alistair Salisbury
Approved Matt Madden Alistair Salisbury Anton
02 2011-12-08
Sluijterman
CONTENTS 3
2.0 INTRODUCTION 7
3.0 SCOPE 8
5.4 Sizing 10
Caissons Offshore 27
10.3.1 Criteria 30
14.0 VIBRATION 38
REFERENCES 39
APPENDIX B: VIBRATION 43
Shortening Meaning
API American Petroleum Institute
BFD Basis for Design
BS British Standard
CGR Condensate to Gas Ratio
CI Corrosion Inhibitor
CITHP Closed-in Tubing Head Pressure
DEP Shell Design and Engineering Practice
DN Nominal Diameter – used in metric pipe sizes (mm)
FID Final Investment Decision
FEED Front End Engineering & Design
GRP Glass-fibre Reinforced Plastic
HIC Hydrogen Induced Cracking
ISO International Organization for Standardization
NACE National Association of Corrosion
NFPA National Fire Protection Association
NPSH Net Positive Suction Head
NPSHA Net Positive Suction Head Available
NPSHR Net Positive Suction Head Required
ORP Opportunity Realisation Process
PEFS Process Engineering Flow Scheme
PFS Process Flow Scheme
PIMS Pipeline Integrity Monitoring System
PR Peng Robinson Equation of State
PVT Pressure Volume Temperature (data for hydrocarbon fluid)
QA/QC Quality Assurance/Quality Control
Z Gas Compressibility Factor
This document should be seen as reflecting current best practice and should be revised regularly
to incorporate learning from design activity, research and technological developments where
appropriate. This document is meant to act as a guide only. Shell standards, such as
DEPs, and regional standards supersede this document and should be consulted during
the design process.
Line size is dependent on required inlet and outlet conditions, namely the allowable pressure
drop, quantity of fluid flowing, and the physical properties of the fluid. There may also be a
maximum allowable velocity to be taken into account for reasons of corrosion and erosion, or
excess noise and vibration.
When sizing lines, the main sizing criteria, in the absence of other factors, should be minimum
life cycle cost (reference 46). This may include evaluation of functional requirements, cost of
piping, weight, environmental, energy costs, mechanical and process limitations, expected
lifetime of piping, maintenance cost, etc. Typically, this evaluation will be performed in a
qualitative and semi-intuitive way for plant pipework using rules enshrined in materials selection
and line sizing rules-of-thumb. Pipelines that are individually significant contributors to capital
cost and schedule, often justify a more refined optimisation based on a life-cycle cost-benefit
analysis.
In many cases, these factors will govern the pipe size selection rather than life cycle cost
evaluation.
The process specification of pipework for a typical greenfield project can represent a significant
part of design man-hours. In order that this effort is expended in an efficient way, it is necessary
to perform the line sizing to an appropriate level of detail at the various phases of the project. It
is also necessary to select sizing methods that are fit-for-purpose, meeting project schedule
without compromising the integrity of the design. In many cases, it is possible to use velocity-
based sizing criteria to efficiently line-size. Expeditious use of simple methods will allow effort
to be focussed on critical lines such as pump suction, inlet and outlet relief valve, vent and
blowdown lines.
The selection of pipework and pipeline materials should be made according the project materials
selection philosophy. The materials selection philosophy is developed by the
Corrosion/Materials Discipline in conjunction with Process and Mechanical Disciplines who
provide information from the heat and mass balances.
It is the responsibility of the Process Engineers to initially select the mechanical design
temperatures and pressures for the process piping. However final design conditions should be
selected in close consultation with the mechanical design engineer. Consideration should be
given to the maximum and minimum temperatures and pressures that could be experienced
during start-up, shutdown, rapid depressuring, and any other relevant transient conditions.
Detailed guidelines for selecting the mechanical design conditions can be found in DEP
01.00.01.30-Gen (Reference 17).
5.4 Sizing
Piping is sized for the controlling operating case determined by analysis of flow rates, operating
pressures and temperatures for all identified operating modes. Consideration should be given to
start-up conditions and off-design operation where these significantly affect line size. The
remainder of this document is dedicated to techniques for sizing the diameter of a line.
Graphical and Tabular methods represent a convenient and effective method for preliminary
liquid line sizing (Reference 1 and 2) based on maximum velocity criteria.
Graphical methods are generally not used for gas or multiphase sizing because the mathematical
relationships are more complex due to the compressible nature of the fluids and the complex
behaviour of multiphase flows. Even so, graphical methods should only be used as a first pass
for line sizing estimates.
It should be recognized that, tables and graphs used for line sizing are generally based on new
pipe. Corrosion, erosion and solids deposition may foul the line over time increasing hydraulic
roughness and the friction factor: a 5% decrease of the internal diameter of the pipe will
increase pressure drop by more than 25% for the same flow.
Manual calculations can be used for the following line sizing situations:
Manual calculation is not recommended when the method is long and complex and therefore
difficult to validate and inefficient; examples are:
Spreadsheets offer a convenient way of performing line sizing. Spreadsheets prepared according
to Reference 5 should preferably be used. Where they are not, the line sizing spreadsheet itself
should be subject to a QA/QC check as appropriate. Both of the tools cited below can be
obtained by contacting the Static Mechanical group in Upstream Major Projects - Americas.
a. API RP 14E Line Size Spreadsheet – This is an Excel workbook used to find pressure
drop & velocity for single phase or multiphase flow of oil, gas, and or water in order to
size pipe. Most formulas in the Excel line sizing worksheet are taken from API RP 14E,
Recommended Practice for Design and Installation of Offshore Production Platform
Piping Systems with the exception of the Churchill Equation which is used to accept all
ranges of Reynolds number Re and e/D to determine the friction factor. The
predominant API RP 14E formulas are modified Darcy-Weisbach and Fanning
equations. Pressure drop is calculated per 100 feet of pipe. The Weymouth, Spitzglass &
Hazen Williams formulas are also used. The spreadsheet is limited to 4 pipe sections & is
used primarily for carbon steel pipe. API 14E is not intended to be used for erosion
velocity calculations.
The main role of a chemical process simulator like Unisim is to generate physical properties and
plant flows for use in line sizing. It is not common practice to use these tools for the purpose of
line sizing itself.
When sourcing physical property data for line sizing, consideration should be given to the
following:
a. The „precision‟ of the simulator‟s physical properties is often not required for line sizing
so consideration should be given to alternative fit-for-purpose methods, such as simple
Z factor correlations (Reference 6), where this will save time.
b. Some of the physical property data from the simulator may not be reliable. This is
particularly true for heavy (i.e. low API gravity) crude lines where the viscosity
correlations may not model the crude behaviour well and where emulsion formation may
be an important factor. Predictions for these systems should be validated wherever
possible against laboratory data.
The PIPESIM program is available on the Shell Computer System. Pipesim is typically used for
flowlines and export lines, but can be used for simulation of lines in plants, which are typically of
complex multiphase flow. Black-oil or compositional models may be used. Generally,
compositional models should be used for dense phase or condensate systems where the CGR
(condensate-gas-ratio) is of the order of 102 bbl/MMSCF or less, as the liquid loading is sensitive
to pressure and temperature. Black-oil models can be used for higher liquid loaded oil
dominated systems.
Flare network simulators can be used for predicting instantaneous pressure and flow rate
information based on specified initial conditions. Two such tools within Shell are Flarenet and
Visual Flow. Among other functionalities, these steady state simulators are capable of modeling
pressure safety valves, blowdown valves, and piping.
Pipe-Flow is a commercial software tool that analyzes steady state hydraulic networks both open
and closed using the Darcy-Weisbach formula and the Colebrook equation to calculate the
friction factor in the pipelines. Pipe-Flo allows users to analyze the interaction of pumps, pipes,
control valves, and other components in a Windows graphical interface. The program supports
all valve and fitting types found in the Crane Technical Paper 410 and allows the addition of
custom valves and fittings.
Generally, dynamic simulation plays a minor, secondary role in line sizing however, it is
frequently used to evaluate the impact of a selected line size on the design. The best example of
this is the sizing of slug catchers at the end of multiphase pipelines. Dynamic simulations are
also sometimes used to aid materials selection based on design conditions developed from an
understanding of the dynamics.
A. OLGA
This is a transient multiphase simulator developed by Scandpower (Reference 43). It is capable
of one dimensional dynamic simulation of oil, gas, and water flows in flow line and pipeline
systems.
B. Compas
This is a Shell developed dynamic pipeline modelling add-in for Unisim. It is capable of one
dimensional dynamic simulation of oil, gas, and water flows in flow line and pipeline systems.
Compas can be used for stand-alone pipeline modelling as well as for integrated modelling with
inclusion of the facilities.
D. Blowdown 2000
This program, developed by Imperial College, can be used to predict temperature and pressure
profiles occurring during vessels and pipelines during depressuring (Reference 44). Blowdown
2000 has the advantage that the results predicted by the program have been validated against real
world data.
For in-plant pipework (i.e. that is normally designed to ASME B31.3 code), outside diameter and
wall thickness (schedule number) of welded and seamless steel pipe is standardised and should
be used with the following limitations:
A minimum size of DN50 (2") should be used for all process, process support and utility piping
to ensure adequate mechanical integrity. Smaller piping can be used, where protection and/or
support is provided to withstand human activity.
If using tubing make sure to use the correct dimension data for the tubing, i.e. ½” tubing is not
equivalent to ½” piping.
Friction factors may be determined graphically using Moody charts such as those given in
Reference 8.
The friction factor may also be calculated using correlations such as that due to Churchill
(Reference 39).
Where
f is the Moody Friction Factor (which is 4 times the Fanning friction factor)
is the absolute roughness in mm
D is the pipe diameter in mm
Re is the Reynolds number (based on the average velocity and on the pipe diameter)
log is the natural logarithm
Note: As long as and D are in the same units, the equation holds.
For all calculations of pressure drop, where the materials are known, the following pipe
roughness values should be used:
Where scaling is expected, an allowance for scale build up should be included. The likely
thickness of scale should be established from field chemistry and experience. In the absence of
firm data this can be taken as 5mm with a roughness of 0.15mm.
For rigorous pressure drop calculations, the factors presented in Reference 7 should be used.
For items not covered in Reference 7, Reference 8 can be used.
For initial estimates, where the number and type of fittings are unknown, the equivalent length
of piping may be obtained by multiplying the piping run obtained from preliminary layout
drawings by the factors given below:
Fluid physical properties are required to size some but not all lines. Lines that do not require
physical properties to be sized fall into the following categories:
1. Lines that are sized by „rule-of-thumb‟. This applies to lines such as some drains and
utility lines that may be sized based on general practice e.g. 3” drain lines from vessels,
instrument bridles, 1½” air lines etc.
2. Liquid lines where the flow rates are defined volumetrically and the sizing criteria are
velocity based.
Lines that do require physical properties are:
1. Liquid density
2. Gas density and compressibility factor (Z)
3. Vapour liquid split and phase envelopes for multiphase pipelines
4. Liquid and emulsion viscosities
5. Gas viscosity
6. Gas and Liquid Specific Heat and Enthalpy
7. Gas and Liquid thermal conductivity (when heat transfer is an issue)
8. Hydrate formation curve
9. Wax/asphaltene deposition conditions, pour points etc.
It is important to appreciate the level of accuracy required from this data for line sizing
purposes. In some cases line size and hydraulics will be very sensitive to fluid properties, in
other cases fluid properties have little or no (see above) influence on the selected line size. In a
typical design there will be hundreds of lines to size and it is important not to waste time
defining physical properties with excessive accuracy where the line size is not sensitive to this
information. Typical examples of this might be gas viscosities for turbulent gas flow or thermal
conductivities for insulated pipe.
Situations where special care is required on ensuring fluid properties data is accurate and valid
are discussed in the following sections.
For Reynolds numbers between 2000 and 4000, the flow lies in the critical zone. In this region,
the flow may be either laminar or turbulent depending on several factors. The friction factor in
this region is indeterminate, bounded by a lower limit given by the laminar flow regime and an
upper limit based on turbulent flow conditions. For Reynolds numbers below 2000, the flow is
laminar and the friction factor given by Poiseuille's Equation:
In this case the friction factor and pressure drop are proportional to viscosity and typically 3 to 4
times higher than in the turbulent flow regime.
Because of this, care should be taken when the viscosity of the crude or its emulsions are of the
order 102 cps or greater. This particularly applies for heavy and waxy crudes where the rheology
can become non-Newtonian (i.e. demonstrating a viscosity that varies with shear rate). By their
nature, heavy crude wells have flowing tubing head pressures and available pressure drops that
are low. This factor combined with the strong dependence of friction factor on viscosity means
that the line sizing is very sensitive to liquid viscosities. This problem is compounded by the
strong dependence of heavy crude viscosity on temperature and emulsion behaviour.
In these situations, viscosity should be derived from measured data wherever possible. Viscosity
derived from process simulation package should be treated with caution.
For sizing calculations of lines carrying oil and water where emulsions are likely to be formed,
due regard should be given to the possible high emulsion viscosity, as this can be higher than the
individual phases on their own. For estimating viscosity to be used in sizing calculations, the
following Guth and Simha equations may be used (References 30, 31):
e w oiw w w wio w w
Where:
[ 1 tanh [ n ( a b ) ] ]
( a b )
2
is a „flip‟ function to simulate the inversion of the emulsion. n, (n>0) is a coefficient that affects
the sharpness of the transition at the inversion point.
A typical curve is shown below with some experimental data from one of the Gannet crudes.
700
600
Emulsion Viscosity cP
500
400
300
200
100
0
0% 20% 40% 60% 80% 100%
Water Cut
This equation has some basis in theory, however, where strong emulsions are anticipated such as
in low API gravity crude, it is recommended that laboratory tests be performed to calibrate the
viscosity predictions.
Another correlation that is frequently used for oil production systems is the Woelflin correlation
However, at higher water cuts (greater than about 40%) it tends to be excessively pessimistic
leading to higher pressure drops than are actually likely to occur (Reference 31). The Shell
Dehydration Manual provides further insight into emulsions and dispersions (Reference 49).
Many of the liquid lines will be sized initially prior to any detailed knowledge of layout. There is
little justification in spending large amounts of time on complex hydraulic calculations. Initial
line sizing should be performed, as far as possible, using graphical methods where velocity is a
limiting factor. When more detailed layout information is available, sizes for critical items such as
pump suction lines should be confirmed using pressure drop calculations.
The sizing criteria for liquid lines will depend on application. In general, the function and
application of the line will determine the sizing criterion to be selected. Where pressure drop is
not a determining parameter, the size should be determined by the velocity constraints given in
Section 9.2. The following applications are not sized according to maximum allowable velocity.
These require special considerations that are described in the following sections.
The velocities should in general be kept low enough to prevent problems with erosion, water
hammer, pressure surges, noise, and vibration and reaction forces. In some cases, a minimum
velocity is required. Table 2 shows recommended maximum velocities for various services and
piping.
Table 2 Maximum Velocities for Sizing Liquid Lines (Norsok, Reference 34)
When the service is intermittent, the velocity can be increased to 10 m/s. For CuNi the
maximum velocity is 6 to 10 m/s depending on duration and frequency of operation.
Figures for approximating minimum and maximum recommended liquid velocities for various
pipe sizes can be found in API RP 14E, Reference 10. It is typically advisable to add a 20-50%
margin or “surge factor” to the anticipated normal flow rate. That is, unless more detailed surge
analysis has been performed.
The suction piping should be sized based on NPSH requirements. The NPSH is the total
suction head in feet of the liquid being pumped minus the absolute vapour pressure in feet of
the liquid being pumped. See the example below for an NPSH calculation. The following
maximum velocity and maximum pressure drop gradient criteria should be satisfied for
preliminary sizing purposes:
The fluid temperature should be at least 15C below the fluid boiling point temperature to allow
the sizing to be based on the criterion for sub-cooled liquids.
Reference 9 specifies the Shell DEP requirements for margins between NPSH required
(NPSHR) and NPSH available (NPSHA).
Open system, source above the pump at 68°F. Atmospheric pressure is 14.696 psia, the vapour
pressure of the liquid is 0.339 psia, and frictional losses are 2.92 ft
Where ha is the absolute pressure (in feet of the liquid being pumped) on the surface of the
liquid supply level, hvap is the head in feet corresponding to the vapour pressure of the liquid at
the temperature being pumped, hst is the static height in feet that the liquid supply level is above
or below the pump certerline, and hfs is the suction line losses (in feet) including the entrance
losses and friction losses through the pipe, valve, fittings, etc.
NPSH = 40.26 ft
Reference 9 also provides special requirements „where liquids contain dissolved gas. It should be
assumed that this refers to situations where a light component is chemically combined with the
liquid. An example of this would be pumps installed to pump rich amine in CO2 removal
systems. The effective Henry‟s Law coefficient is lower than for the case of physical solution on
its own. Consequently, the volume of gas liberated as the fluid accelerates into the eye of the
impeller is higher than would be experienced during pump NSPHR testing by the vendor that is
normally performed with water.
For reciprocating pumps, the suction piping should be based on NPSH requirements. The
following guideline should be considered when specifying discharge piping to minimize
pulsations:
These limits are for a single plunger pump without installed pulsation dampers. The fluid
velocity from a reciprocating pump varies over time – the velocities above are time-averaged.
Increasing the number of plungers and/or installing pulsation dampers may be used to increase
the velocity limits in the table above (Reference 21, 22). Reference 22 includes an example of
how to size a pulsation dampener.
Allowance should be made for the acceleration losses in reciprocating pump suction lines. For
simplex pumps, the suction and discharge lines should be sized for 1.6 times the pump rate. For
other reciprocating pump configurations, follow references 1 (Section 5.3.2.4) and 10.
Numerous difficulties have arisen on production separator outlet liquid flowmeters where gas
break-out has occurred. In many ways this is analogous to the pump NPSH problem. In some
cases these have been sufficiently severe to render the flow meter in-operable and mitigation can
be very difficult (if for example a vessel elevation change is required).
In order to avoid this issue the total fluid pressure needs to be sufficiently above the vapour
pressure to avoid gas breakout. Unless a pump is added (which for custody transfer is often the
case), static head due to relative elevation. The following points need to be borne in mind :
The static head (elevation) from lowest liquid level down to the flowmeter inlet needs to
be sufficient to account for nozzle losses, piping friction and unrecovered meter losses.
Inlet piping should be designed without high points sufficient vapour break out as it is
unlikely that recombination will occur even if the flowmeter elevation is correct.
It is preferable to design (liquid) control valve inlets for liquid service to avoid gas breakout in a
similar fashion to the flowmeter discussion to avoid capacity constraints due to the vapour
phase. Where this is not feasible, and resultant vapour would have a significant impact on
control valve sizing and specification, the control valve sizing and specification basis needs to be
adjusted
Gravity flow lines are those where the liquid flows by gravity forces instead of pressure
difference. This includes tank overflows, drains (sanitary, closed and open drains).
Problems have occurred on facilities with significant elevation changes, such as multi-deck Oil
and Gas platforms, associated with routing „gravity‟ flow pipes such as overboard water lines.
Problems of pulsation and vibration due to formation of vacuum if inadequate vacuum breakers
are provided (causing cavitation for example), or inadequate consideration of air entrainment
where vacuum breakers are installed, can occur in this piping as well as the overboard caisson
itself. As an example, a water pipe transporting water downhill to an atmospheric sump, would
likely have sub-atmospheric pressures upstream.
The size of tank overflow lines should be greater than or equal to the size of the largest inlet
pipe as a minimum starting point. However particularly where the overflow line is routed a
significant distance from the tank, a rigorous hydraulic analysis is needed
Generally, for fixed installations, a minimum downward slope of 1:100 should be used.
However, with mud and/or sand, the slope should be at least 1:75.
On floating installations, the slopes must be evaluated according to planned installation trim.
Drain piping should be sized (running full) for the maximum expected flow. This rate should be
based on the requirements to dispose of firewater, rainwater and spills. Guidelines for
determining these rates are contained in Reference 35.
Fluid velocities in drainage piping should be limited to a maximum of 0.9 m/s. This velocity will
generally prevent dispersion of oil by turbulence. This velocity may be exceeded under the fire
fighting design condition.
Higher velocities are also permitted in systems that are primarily designed for de-ballasting
systems and cooling water networks, to avoid uneconomical pipe sizes. Designs should aim for
uniform stable flows (e.g. large radius bends, no tee junctions etc.) Velocities of less than
0.3 m/s should be avoided in order to prevent a premature separation and consequent
Pipes that are running full, and do not require a minimum downward slope to avoid particle
deposition, should be sized according to the total available static pressure head, and the
maximum allowable velocities for liquid lines.
If operated at atmospheric pressure, the design pressure of the piping should be 10 barg. This is
based on the requirement to allow the connection and use of fire hoses for flushing and cleaning
(References 23 Section 4.3.3, 24). The design pressure of collection vessels should be 3.5 barg
minimum.
The minimum size for the sewage and open drain header should be DN100 (4") and sub-headers
DN80 (3").
Vertical sections of line flowing vertically downward, such as discharge caissons should be self-
venting and avoid air entrainment. In order to achieve this, these lines are sized to keep the
Froude Number <0.3 (Reference 31). The following formula should be used for sizing:
0.4
d
4 Q
0.5
( 0.3) ( 3600) g
Where:
d = pipe ID (m)
Q = liquid volumetric flow (m3/h)
g = acceleration due to gravity = 9.81 m/s
If it is not possible to design for self-venting flow, the piping should be sized to give a Froude
number greater than 1.0 so that
0.4
d
4 Q
0.5
( 1.0) ( 3600) g
Note that the above equations should be used as there are number of different definitions of
Froude number that sometime causes confusion.
A drainage box is a collection system that may be considered a vessel or an atmospheric sump
where the deluge water flow is considered to be larger than the rainfall amount. Drainage of
deluge water from drain boxes through vertical lines should be sized on the basis of 50% of the
available head (assuming the pipe running full of liquid) and not Froude number. The following
method can be used to determine capacity:
2.5
8855.8d
Q
f
Q = flow in m3/h
f = Moody friction factor
D = pipe ID (m)
1
f
2
2 log d 1.74
2
The formula is derived by equating 50% of the available hydrostatic head with the friction losses
in the pipe loss:
d g
V max
f
Where Vmax is the maximum velocity in m/s and g is the acceleration due to gravity
A vent line is normally included from top of the vertical gravity line from platform topside to sea
for seawater and produced water discharge lines. The vent line should be designed for an air
volumetric flow rate equal to the liquid volumetric flow through the vertical line and a pressure
loss of maximum 0.02 bar/100 m.
Fire water system line sizing should be performed in close consultation with Safety Engineering
Specialists. Line sizing is required to ensure the correct delivery of flow to deluge nozzles,
monitors and hose stations at various locations in the plant. The line sizing of fire water lines
should be based on supply and delivery pressures and allowable flow velocities. The pressure
drop to large deluge systems should be calculated based on the most hydraulically unfavourable
pipe routing. In the ring-main pipework, the flow velocity should not exceed the values
specified in Reference 25 Section 4.1.4.
For offshore systems only, upstream of the deluge skids, the flow velocities should not exceed
10 m/s. Some areas may require velocities higher than 10 m/s in order to balance systems
hydraulically. This is acceptable if the pipework is adequately supported to accommodate
reaction forces and prevent flow-induced vibration. The design should be performed in
accordance with Reference 36, Section 3.2.3.8
Pressure drop calculations through fire fighting hose and fittings should be performed using
vendor data e.g. Reference 11.
In many cases, the design of firewater systems requires the development of a detailed network
model by specialist contractors
The lines for oily-water to water treatment systems should be sized in order to retain the size of
oil droplets in the water. This can be achieved by providing low velocities. Typically, the
velocity should not exceed 3 m/s. This should be considered in the selection of fittings and
instruments in these lines to avoid sharing of oil droplets.
The minimum flowing velocity of drilling fluid should not be lower than 0.8 m/s in order to
prevent the settling of sand in pipes. The maximum velocity in carbon steel should not exceed
4 m/s to avoid problems such as cavitation/erosion on bends and damage to inline equipment
and vessel internals. The maximum pressure drop gradient in pump suction and gravity flow
pipes (carbon steel) should be 0.3 bar/100m.
The following method has been used successfully to size disposal caissons for produced water,
cooling water, fire water etc where the following conditions are met :
In practice, for practical services and applications offshore, these conditions can be met. Indeed
this type of configuration may not be suitable at all if these conditions cannot be met. The basis
of the sizing is to ensure a stable annular flow of water around a central core of air, which then
prevents violent „slugging‟ and pressure surges. The correlations and sizing criteria were
developed for use in the civil engineering (plumbing) industry for drainage stacks from
skyscrapers and other tall structures (References 51 and 52). The criteria for stable flow used in
this industry is less than 25-50 mm W.C of internal vacuum or pressure fluctuation to avoid
disruption to upstream drainage sources.
Where α (ft2 ) is cross sectional area assumed full of liquid per the criteria above
d is internal diameter (ft)
Q is the volumetric flow rate (ft3/sec)
This section covers normal process lines but excludes flare lines where higher velocities are
allowed.
It should also be recognized that high-pressure drops are associated with increased noise,
vibration and erosion.
There are generally no minimum velocity criteria for lines handling single-phase gas above its
dew point. In certain situations a minimum flowing velocity should be considered in the sizing
process:
a. Where there is the potential for sand or solids to prevent dropout in pockets in lines.
b. Where film-forming corrosion inhibition is used, a minimum velocity is required to
ensure that the corrosion inhibitor coats the pipe wall. This minimum velocity is
typically ~5 m/s. This should be confirmed with the chemical vendor and by multi-
phase flow pattern analysis to ensure that the flow is in the annular-mist or mist flow
regime.
c. Where significant levels of liquid carry-over are possible, the lines should be sloped to
allow free drainage to collecting vessels. Where this is not possible a minimum velocity
will be required to ensure the liquid does not collect in low points (generally this
situation should be avoided).
d. Where chemical injection is required to be sprayed into the gas phase, a minimum
velocity may be necessary to ensure adequate mixing. An example of this would be
H2S scavenging by chemical injection.
The maximum velocity criteria described here apply to general process and utility lines. They do
not apply to flare and vent lines that are subject to infrequent and short-term use and can be
designed for higher velocities. The maximum continuous velocity should not exceed erosional
velocity limits specified in Section 12.1.
The limits for „air and other gases‟ above are general and conservative and the potential for
employing higher velocities should be considered where justified by life-time cycle cost
reduction and system integrity is not compromised by, for instance, unacceptable levels of noise
and vibration.
10.3.1 Criteria
When pressure drop is critical (e.g. when it results in unacceptable liquid drop out in suction
lines between scrubber and compressor suction) pressure gradients shown in Figure 2 should be
used.
Criteria for Sizing Gas Lines Based on Pressure Gradient (Norsok Reference 34)
Operating Pressure (barg) Pressure Drop (bar/100 m)
0 to 35 0.001 to 0.11
35 to 138 0.11 to 0.27
Over 138 P/500 [1]
The following equation, which is found in Reference 20 Section 2.3.3.2, provides a formula for
pressure drop for liquid service. For approximation of pressure drop in gas service, this equation
can be used with the following conditions
1) If pressure drop is less than 10% of the upstream pressure, use density and velocity
based on either the inlet or outlet conditions.
2) If the pressure drop is between 10% and 40% of the upstream pressure, use density and
velocity as averages of inlet and outlet conditions.
Where ΔP is the pressure drop in Pa (1 Pa = 1 N/m2, 105 Pa = 1 bar, 1Pa/m = 10-3 bar/100m)
λ is the friction factor (which depends on the Reynolds number and the roughness factor)
D is the ID of the pipe in m
is the density in kg/m3
L is the total design length (m)
V is the average linear flow velocity (m/s)
Erosion is an issue particularly for flowlines conveying well stream fluids that frequently contain
suspended solids. Mechanical damage can result from unbalanced forces that exist in multiphase
flows due to fluctuations of fluid density and velocity. These unbalanced forces are very
effective at exciting pipework creating low frequency (of the order of the natural frequency of
the pipe span), high amplitude vibrations with concomitant stress cycling. This can rapidly lead
to fatigue failure of attached small bore connections and even the pipe itself. These effects are
particularly acute for the intermittent (i.e. slug or plug) flow regime and steps should be taken to
avoid this whenever possible.
If the available pressure drop allows, the velocity should in general be sufficiently high to ensure
a well mixed flow (i.e. mist flow or bubbly flow). This prevents instabilities due to liquid
accumulations, and it allows simple pressure drop calculations. If lower velocities are required
due to limited available pressure drop or at turndown situations, problems with slugging and/or
liquid accumulation in the lines should be considered.
The process engineer should include notes on the PEFS and line-list to alert the piping engineer
to the presence of multiphase flow so that pipework and associated small- bore connections are
adequately supported and reinforced. For a detailed discussion of design and operation of multi-
phase lines see Reference 47.
Unbalanced forces arising from fluctuating fluid density are generally calculated from the
maximum kinetic energy density of the flow. For non-slug flow, this parameter is evaluated
from homogeneous densities and mixture velocities i.e.
wg ww wo
m
wg ww wo
g w o
wg ww wo 1
vm
g w o Ap
Where m is the no-slip mixture density, wg, ww, wo are the mass flow rates of gas, water and oil,
g, w, o are the densities of gas, water and oil, Ap is the cross-sectional area of the pipe.
If mvm2 > 200,000 Pa the piping discipline should be consulted in order to consider reaction
forces.
The simplest approach is to assume the liquid density for and the superficial gas velocity for v.
For a typical flowline, 800 < < 1000 kg/m3, giving a maximum velocity of 14 to 15.8 m/s.
However, in practice, liquid slugs are not sustainable at higher velocities and the slug density is
considerably less than that of single-phase liquid. Application of this approach, particularly in
flare systems may lead to over design by trying to design to accommodate slugs of dense liquid
travelling at Mach 0.5! This is further discussed in Section 13 on the sizing of Flare Lines.
Prior to this software availability, API 14E was used for erosion calculations. However this
method has been shown to be both under and over conservative depending on the solids
loading. The API 14E method should not be used for erosion calculations.
While many existing facilities have been designed using the API 14E approach, this method is
outlined for reference only. The API 14E approach should not be used for new designs.
The maximum velocity to limit erosion is calculated by the API RP 14E formula (below).
C
V c
m
Where
Both ISO 13703 and API RP 14 E recommend the following values for C:
For carbon steel piping systems, the corrosion rate often limits the lifetime. With increased flow
velocity, the corrosion rate tends to increase due to increased shear forces and increased mass
transfer.
For initial sizing, the flow velocity should be restricted to maximum 20 m/s to limit the erosion
of the protective layer of corrosion products and reduce the risk for a corrosion inhibitor film
break down. This velocity limit is for stripping inhibitors; 20 m/s should be achievable with all
modern inhibitors. Inhibitors that work at higher velocities are available, but these require
specific testing. A suitable sizing velocity should be determined in conjunction with the
corrosion engineers as part of the project corrosion prevention philosophy.
Where it is proposed to protect the pipework or pipelines with film-forming corrosion inhibitor
the sizing may be based on a velocity that maintains the integrity of the corrosion inhibitor on
the pipe wall. For effective distribution, the pipeline velocity needs to be maintained high
enough to be in the annular-mist flow regime. However, if the velocity is too high, the film is
removed from the pipe wall. It is recommended, if such inhibition systems are considered, that
discussions with chemical vendors are conducted early in design and, if necessary, laboratory
tests performed to establish an acceptable velocity range. These parameters can have a
significant effect on pipeline specification.
The minimum velocity should be verified in order to avoid the breakdown of protective scale by
solids deposited from the pipeline.
Starting in the Define Phase and continuing into the Execute Phase a detailed model of the flare
system network should be developed using a specialist computer program such as Flarenet or
equivalent. An analysis should be performed for all controlling relieving, depressuring and
process flaring cases to ensure that the maximum allowable backpressure at each individual
relieving device is not exceeded.
Although this guideline is not concerned with material section, it should be noted that
conditions in flare systems often require low temperature materials.
Relief valve inlet line sizing is generally performed during the project „Execute‟ phase. During
the „Define‟ Phase, relief valve and associated pipework sizing has not been performed and
hence the size should be identified by a „HOLD‟ on the PEFS and line list. The following points
are meant to serve as rules-of-thumb for inlet line sizing:
a. Relief valve inlet line pressure drops should be less than 3% set pressure based on a
maximum flow through the installed relief area. Inlet pressure losses that exceed this
value cause valve chatter that has the potential to
i. Cause severe damage to the valve and surrounding pipework caused by severe
vibration;
ii. Reduce the effective relief area, as, on average, a chattering relief valve is only
partially open.
b. For initial estimates, an equivalent length of 40m can be assumed in the absence of
specific piping layouts. The pipe absolute roughness selected should be as follows:
c. As a rule-of-thumb, the inlet line should also be at least one nominal size greater than
the relief valve inlet flange. Some small low capacity relief valves, such as those
installed for thermal relief, may not need the increased size. However, in all cases, the inlet
pipe bore area should be greater than the installed area of the relief valves.
d. Relief valve inlet lines should be self draining and un-pocketed. This requirement
should be clearly identified on the PEFS.
See Reference 13 for detailed guidelines on sizing Relief valve discharge lines. The following
points are meant to serve as rules-of-thumb for discharge line sizing:
Guidelines for acceptable Mach number in headers and sub headers can be found in Reference
14. Where these criteria cannot be met, additional calculations should be shown to document
that the selected pipe size is still acceptable. This involves evaluating piping stress levels, pipe
supports, noise, etc. The v2 value should not exceed 200,000 Pa. This criterion provides a
shorthand check for vibration, fatigue, noise, and erosion (Reference 34)
Controlled flaring lines downstream of pressure control valves should be designed for a
maximum velocity of Mach 0.5.
In the lines, upstream or downstream of the blowdown valve, the value of v2 should not exceed
200,000 Pa.
Where large inventories of cold gas are being depressurised, an assessment should be made of
the effect of pipe chilling upstream and downstream of the blowdown valve. The pipework
downstream of the blowdown or depressuring valve needs to be designed for low temperatures
Tail pipes downstream of relief valves designed for full flow relief should be designed to
accommodate high velocity two-phase flow. Furthermore, the flare system pipework is often
fabricated from Schedule 10S 316 Austenitic Stainless Steel. This makes the pipework
potentially very flexible and vulnerable to damage from large forces generated by the unbalanced
flow of fluids. In order to assess the severity of these loads it is necessary to calculate the peak
v2 for the flow.
Multi phase relief lines should be sized based on the following criteria:
Note that for case (a), slug or plug flow are unlikely to exist at velocities approaching 50m/s in a
steady state situation. They could exist transiently in certain situations such as in the pipework
immediately downstream of bursting discs on liquid filled system.
The recommended approach for calculating v2 is to model the tail pipe flow with Flarenet or
similar program and calculate the maximum value of v2 based on the assumption that the flow
is homogeneous. In some cases, such as bends downstream of bursting discs protecting shell-
and-tube heat exchangers, a high velocity transient liquid slug may be a credible scenario.
Pipework that have v2 approaching 200,000 Pa should be checked using the methodology in
Reference 27 Appendix A2
If v2>200,000 Pa mitigation measures should be taken in consultation with the piping stress
engineer. Typical examples are:
Reaction forces generated by the lifting of relief valves need to be accommodated by adequately
supporting the valve and associated pipework. The formulae for calculating the reaction forces
are contained in Reference 38.
Special consideration needs to be given to sizing atmospheric flare and vent systems that operate
at a few milli-barg. These are frequently connected to tanks that have design pressures of 20
mbarg for „Low Pressure‟ Tanks and 56 mbarg for „High Pressure‟ Tanks. The venting
requirements for these tanks are described in Reference 42. The pressure drop may be
calculated using the ESI program or the methods outlined in ISO 13703 (Reference 1) Section
5.4.2. Note that the pressure drop is low relative to absolute pressure hence gas density changes
along the pipe will be small allowing the use of the incompressible flow equation with
compressibility factor, Z=1.
Where atmospheric vents are attached to vertical caissons on offshore platforms, account should
be taken of the effect of the pumping action of the sea on the hydraulics of the vent system.
The main source of noise in process pipework is generated across valves. The piping
downstream of the valves is considered the principal radiator of the generated noise. There are
three main sources of valve noise:
a) Mechanical Vibration
Mechanical noise is caused by the response of internal components within a valve to turbulent
flow through the valve. Vortex shedding and turbulent flow impinging on components of the
valve can induce vibration against neighbouring surfaces. Noise generated by this type of
vibration has a tonal characteristic.
b) Aerodynamic Noise
Aerodynamic noise is a direct result of the conversion of the mechanical energy of the flow into
acoustic energy as the fluid passes through a valve restriction. The proportionality of conversion
is called acoustical efficiency and is related to valve pressure ratio and design.
c) Hydrodynamic Noise
Liquid flow noise, cavitation noise, and flashing noise can be generated by the flow of a liquid
through a valve and piping system. Of the three noise sources, cavitation is the most serious
because noise produced in this manner can be a sign that damage is occurring at some point in
the valve or piping. Unlike aerodynamic noise, hydrodynamic noise can be destructive even at
low levels, thus requiring additional limitations for good valve application practice.\
Typically detailed vibration analysis is carried out by mechanical engineers. A summary of this
analysis is shown in Appendix C and in Reference 20.
1. Petroleum and natural gas industries – Design and installation of piping systems on
offshore production platforms, BS EN ISO 13703:2001, BSI 13th August 2002.
2. Piping General Requirements, DEP 31.38.01.11-Gen. February 2011, Appendices 1
and 2
3. API STD 521 „Pressure-Relieving and Depressuring Systems‟, 5th Edition, January
2007 Section 7.3.1.3, Design of Relief Device Discharging Piping’
4. GPSA Handbook, Section 17, Fluid Flow and Piping
5. Standard Calculation Spreadsheet (SCS) Quality Control Procedure, GEN-EPA-
E14-00013, March 2005.
6. GPSA Engineering Data Book, Section 23, Physical Properties
7. Piping General Requirements, DEP 31.38.01.11-Gen. February 2011, Section 2.3.3.1
8. Flow Fluids Through Valves, Fittings and Pipe, Crane Publication 410M, Section 2.8
and Appendix A-23 to 29
9. Centrifugal pumps (amendments/supplements to ISO 13709) DEP 31.29.02.30-Gen,
February 2011,, Parts III (Section 6.1.8 and 8.3.4.3) and IV (Section 6.1.8)
10. API RP 14E, Section 2.3, Sizing Criteria for Liquid Lines
11. Angus Fire web site www.angusfire.co.uk
12. Selection of Materials for Life Cycle Performance (Upstream Facilities) – Materials
Selection Process, DEP 39.01.10.11-Gen, Appendix D.2 and D.3, February 2011
13. Design of Pressure Relief, Flare and Vent Systems, DEP 80.45.10.10-Gen., Section
3, February 2011.
14. Design of Pressure Relief, Flare and Vent Systems, DEP 80.45.10.10-Gen., Section
3.2, February 2011.
15. GAP Multiphase Network Optimisation Program, www.petroleumexperts.com
16. Two Phase Flow in Pipes by J.P. Brill and H Dale Beggs, Third Edition, Dec 1978 p
4-54ff
17. Definition of Temperature, Pressure and Toxicity Levels, DEP 01.00.01.30-Gen.
February 2011
18. Piping classes - exploration and production, DEP 31.38.01.15-Gen, February 2011
19. Gbaran Ubie, Piping Materials and Design Conditions, GBU-EPA-GBA-G26-
00006-A21, December 2004
20. Piping General Requirements, DEP 31.38.01.11-Gen. February 2011
21. API RP 14E, Section 2.3b
22. See Appendix B for an example sizing calculation and web sites such as
http://www.primaryfluid.com/pdf/Accupulse_PDF/AccuPulse-
Dampener_Sizing_Guide.pdf
23. Drain Systems for Offshore Installations, DEP 37.14.10.10-Gen., February 2011
24. Drainage and Primary Treatment Facilities, DEP 34.14.20.31-Gen, February 2011
25. Active fire protection systems and equipment for onshore facilities, DEP
80.47.10.31-Gen, , February 2011, Section 4.1.4
26. Drainage and Primary Treatment Facilities DEP 34.14.20.31-Gen, , February 2011,
Section 2.4.5
27. Guidelines for the Avoidance of Vibration Induced Fatigue in Process Pipework,
Publication 99/100, Marine Technology Directorate Ltd, 1999.
28. Noise Control Manual, Masoneilan Bulletin OZ3000 01/02,
http://www.masoneilan.com/internet/pages/documentwarehouse
29. Piping Handbook (7th Edition) Edited by: Nayyar, Mohinder L. © 2000; McGraw-
Hill, Page B446
P ulsation damper is a 690 bar 0.5 litre F awcett C hristie dev ice
P umps are Gaso P umps Inc Triplex P lunger P umps M odel 3211-HF w ith 7/8" plunger w ith a capacity of 772
litres per hour. A ssume that plunger is 7/8" x 2" as per dw g B-151-865 in Ref 7. F rom Reference 6:
VP
DP2 LP m3 v olume of plunger
4
Pmin P ( 1 D)
Pmax P ( 1 D)
Pmin 266.2
Pmax 279.8
n
P
1000 VP K
PDQ
Pmin litres required capacity for pulsation dampener
n
P
Pmax
1
Depending upon the type of service, the flow regime, and the equipment being used, significant
vibration can be generated in areas of the process.
If this criteria, where “D” is the inside diameter of the dead leg breach in mm, is not met then
see Reference 20, Appendix 17.2
The DEP, Piping General Requirements, (Reference 7 Sections 3.9) provides guidance on the
design of small-bore connections. The general approach should be to minimise or eliminate
small bore connections in process piping design.
The following guidelines should be considered for optimum piping design (see Appendix of
Reference 28):
i) Straight lengths
Valves should be installed with straight pipe for at least 10D upstream and 20D downstream of a
valve
Isolating block valves, if required, should be selected to ensure minimum resistance to flow. Full
bore type is preferred.
Fluid flow may create higher noise levels than generated by the valve, however this is generally
only the case for gas velocities greater than Mach 0.33 i.e. typically in excess of 100 m/s. In
E&P applications, these velocities are normally restricted to flare systems and consideration
should be given to this source of noise when sizing pipework for valves that are required to
discharge to flare for extended periods.
Expanders and reducers are a source of turbulence in the fluid stream and may be the source of
additional noise. Concentric reducers and expanders with included angles smaller than 30
upstream and 15 downstream of a valve are recommended.
As an exception to the above, short reducers (large included angles) are recommended with LO-
DB restrictors because of their inherent stiffness and the fact that velocity is low upstream of the
restrictors.
Disruptions in the fluid stream, particularly in cases of high fluid velocity, are potential noise
sources. General preferences to avoid turbulence are:
closely coupled
changes in
direction'
'conventional branch'
Piping Supports
A vibration free piping system is not always possible to obtain, especially when using thin-wall
piping such as schedule 5S and 10S. Supports in strategic locations will alleviate many of the
potential structural problems. Thin wall schedule that is subject to mechanical excitation should
be clearly identified with notes on PEFS and line lists so that the appropriate measures can be
taken to design pipe supports.
Pipework downstream of high capacity valves, relief valves, or restriction orifices with high-
pressure drops and corresponding noise generation levels should be checked for acoustically
induced vibration (reference 29). Acoustically induced vibration can cause rapid failure of plant
pipework due to the rapid stress cycling of components that can occur at frequencies of the
order kilohertz. The sections of pipework most likely to be affected by this are the discharges of
relief and blow down valves and the piping downstream of compressor-recycle valves.
Empirical methods have been developed to predict the probability of failure of components
subject to acoustic vibration (References 27 Appendix A2.3, 31).
Where acoustic vibration is identified as a problem, mitigation measures should be considered
including: